Skip to Main content Skip to Navigation
Theses

Estimation des rendements, des besoins et consommations en eau du maïs dans le sud-ouest de la France : apport de la télédétection à hautes résolutions spatiale et temporelle

Abstract : This Ph.D. thesis is part of the MAISEO project associating partners among them: the CACG, managing the water supply of several watersheds located in the south west of France, the Meteo-France center and the CESBIO. One of the goals is to develop innovative and operational tools to estimate crops' water needs at the territory scale. The aim is to provide managers tools to better manage the water supplies linked to the predominant crop encountered in south west of France: maize. The objective of the thesis was to estimate the yield and water requirements of maize crop over large areas. For this purpose, we used an agro-meteorological model coupled to optical satellite imagery. Numerous high spatial and temporal resolution images from different sensors have been used, prefiguring the arrival of the Sentinel-2 data launched in 2015. The first part was to combine remote sensing data with the SAFY (Simple Algorithm For Yield estimates) crop model (Duchemin et al., 2008a) that simulates plant development based on Monteith theory (Monteith, 1972) in order to accurately estimate maize biomass and yield. Numerous field data have been used for the validation at local scale. At regional scale, the results have been aggregated and compared to Agreste yield statistics provided by the French government. Results led us to propose a new formulation of the SAFY model taking into account the temporal variation of the effective light use efficiency (ELUE) and of the specific leaf area (SLA). This modification allows a better simulation of the crop growth dynamics and an improvement of yield estimates at the local and regional scale. Furthermore, we changed the calibration method in order to limit the use of in situ data that are difficult to access over large areas. We also highlighted the contribution of the double logistic function, used to interpolate the NDVI time series. This interpolation enables an accurate determination of the crop growing season and it allows constraining some model parameters such as the emergence date. The SAFY model constrained by remote sensing data is able to well reproduce the yield for the two departments without taking into account the evolution of the soil water storage (Battude et al., 2016)
Document type :
Theses
Complete list of metadatas

Cited literature [271 references]  Display  Hide  Download

https://tel.archives-ouvertes.fr/tel-01804084
Contributor : Abes Star :  Contact
Submitted on : Friday, June 8, 2018 - 3:57:07 PM
Last modification on : Monday, October 19, 2020 - 10:54:11 AM
Long-term archiving on: : Sunday, September 9, 2018 - 6:09:37 PM

File

2017TOU30037B.pdf
Version validated by the jury (STAR)

Identifiers

  • HAL Id : tel-01804084, version 2

Collections

Citation

Marjorie Battude. Estimation des rendements, des besoins et consommations en eau du maïs dans le sud-ouest de la France : apport de la télédétection à hautes résolutions spatiale et temporelle. Sciences agricoles. Université Paul Sabatier - Toulouse III, 2017. Français. ⟨NNT : 2017TOU30037⟩. ⟨tel-01804084v2⟩

Share

Metrics

Record views

197

Files downloads

2599